INDEX

Explanations

specific numbers and data points related to various categories such as politics, health, and technology

oai_token-act-pair · gpt-3.5-turbo Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With GPT2-SMALL @ 8-res_fs3072-jb

Configuration

jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_3072

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

Skylion007/openwebtext

Features

3,072

Data Type

torch.float32

Hook Point

blocks.8.hook_resid_pre

Architecture

standard

Context Size

128

Dataset

Skylion007/openwebtext

Hook Point Layer

Activation Function

relu

Embeds

PlotsExplanationShow Test FieldDefault Test Text

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

NEO

-0.65

 embassies

-0.60

 Ragnar

-0.57

 Icelandic

-0.56

 Authorization

-0.55

 contrasted

-0.55

 Mongo

-0.55

 Hawai

-0.54

 broker

-0.54

 contagious

-0.53

POSITIVE LOGITS

 versa

1.08

 thereof

1.04

 ones

1.01

.?

0.90

.--

0.82

():

0.79

().

0.79

+.

0.77

0.76

 thereto

0.75

Activations Density 2.741%

specific numbers and data points related to various categories such as politics, health, and technology

No Comments

No Known Activations